Cognitive Robotics and Complex Self-Organising Systems (SwarmLab)
What do social networks, biological organisms, economic systems, and the human brain have in common? Their intelligence is in their network: these complex self-organizing systems are more than just the sum of their parts. At DKE SwarmLab we believe that understanding and exploiting the mechanisms that lead to self-organization in complex systems will lead to advanced artificial intelligent systems. One of our key research targets is the creation of cognitive robotic systems – human-friendly machines capable of perceiving, understanding, and interacting with their environment.
In the real world, natural self-organising systems like those mentioned above tend to outperform artificially designed machines. Interdisciplinary research within DKE SwarmLab aims to study self-organising systems to learn their underlying mechanisms. Exploiting this knowledge will allow us to build better artificial intelligent systems that emerge through self-organisation.
Through DKE SwarmLab, DKE holds a chair in the steering committee of the FieldLab Robotics: a consortium of robotics and logistics companies in Limburg. SwarmLab and its partners attracted over 1.5 million euros in research funding since 2015.
Research focus and application
One of the most fascinating applications of complex self-organizing systems are cognitive robots – artificial machines that gain an understanding of their complex surroundings and of the other intelligent entities that surround them. Cognitive robots are capable of planning actions, correctly predicting the results of those actions, executing them, and adapting their behaviour based on their observations. At DKE SwarmLab we focus on the creation of assistive systems – supporting humans in their daily and professional life.
SwarmLab co-develops software and hardware for cognitive robots and complex self-organizing artificial intelligence. Currently running projects involve:
- Intelligent game and assessment devices with AI that assesses well-being, cognitive, physical, and social capabilities of children and elderly.
- Quadcopters capable of autonomously mapping, searching, and navigating in complex environments.
- Self-organization among autonomous agents in nautical processes in modern seaports.
- Smart orthotic devices.
- Advanced brain-robot interfaces.
- EDMO – Educational Modular Robot for teaching and exploring robotics, AI, and data science with students and children.

Scope
Together with our partners, DKE SwarmLab combines interdisciplinary research from a variety of individual research fields, including:
- Cognitive Robotics
- Machine Learning
- Multi-Agent Systems
- Serious Games
- Human-Machine Interaction
- Optimization and Artificial Evolution
- Computer Vision
- Control
- Natural Language Processing
- Assistive Robotics
- Educational Robotics
Go to SwarmLab website
Researchers

Lucas Dahl
- PhD candidate

Yiyong Gou
- PhD candidate

Kirill Tumanov
- PhD candidate
Highlighted publications
- Sprowitz, A. T., Tuleu, A., Ajallooeian, M., Vespignani, M., Mockel, R., Eckert, P., D'Haene, M., Degrave, J., Nordmann, A., Schrauwen, B., Steil, J., & Ijspeert, A. J. (2018). Oncilla Robot: A Versatile Open-Source Quadruped Research Robot With Compliant Pantograph Legs. Frontiers in Robotics and AI, 5, [67]. https://doi.org/10.3389/frobt.2018.00067More information about this publication
- Schmitt, F., Mariyam Christopher, S., Tumanov, K., Weiss, G., & Möckel, R. (2018). Evaluating the adoption of Physical Board Game Ludo as Serious Game for Assessing Cognitive Abilities of Players. In 4th Joint International Conference on Serious Games (pp. 30-42)More information about this publication
- Franz, F., Paredis, J., & Möckel, R. (2017). On the Combination of Coevolution and Novelty Search. In IEEE Congress on Evolutionary Computation (pp. 201-208). IEEE.More information about this publication
DKE research themes
- Affective & Visual Computing Lab
- Algorithms, complexity and optimization
- Cognitive Robotics and Complex Self-Organising Systems (SwarmLab)